Alzheimer's disease (AD), first described by the Bavarian psychiatrist Alois Alzheimer in 1907, is a progressive neurological disorder that begins with short term memory loss and proceeds to disorientation, impairment of judgment and reasoning and, ultimately, dementia. The course of the disease usually leads to death in a severely debilitated, immobile state between four and 12 years after onset. AD has been estimated to afflict 5 to 11 percent of the population over age 65 and as much as 47 percent of the population over age 85. The societal cost for managing AD is upwards of 80 billion dollars annually, primarily due to the extensive custodial care required for AD patients. Moreover, as adults born during the population boom of the 1940's and 1950's approach the age when AD becomes more prevalent, the control and treatment of AD will become an even more significant health care problem. Currently, there is no treatment that significantly retards the progression of the disease. For reviews on AD, see Selkoe, D. J. Sci Amer., November 1991, pp. 68-78; and Yankner, B. A. et al. (1991) N. Eng. J Med. 325:1849-1857.
It has recently been reported (Games et al. (1995) Nature 373:523-527) that an Alzheimer-type neuropathology has been created in transgenic mice. The transgenic mice express high levels of human mutant amyloid precursor protein and progressively develop many of the pathological conditions associated with AD.
Pathologically, AD is characterized by the presence of distinctive lesions in the victim's brain. These brain lesions include abnormal intracellular filaments called neurofibrillary tangles (NTFs) and extracellular deposits of amyloidogenic proteins in senile, or amyloid, plaques. Amyloid deposits are also present in the walls of cerebral blood vessels of AD patients. The major protein constituent of amyloid plaques has been identified as a 4 kilodalton peptide called .beta.-amyloid peptide (.beta.-AP)(Glenner, G. G. and Wong, C. W. (1984) Biochem. Biophys. Res. Commun. 120:885-890; Masters, C. et al. (1985) Proc. Natl. Acad. Sci. USA 82:4245-4249). Diffuse deposits of .beta.-AP are frequently observed in normal adult brains, whereas AD brain tissue is characterized by more compacted, dense-core .beta.-amyloid plaques. (See e.g., Davies, L. et al. (1988) Neurology 38:1688-1693) These observations suggest that .beta.-AP deposition precedes, and contributes to, the destruction of neurons that occurs in AD. In further support of a direct pathogenic role for .beta.-AP, .beta.-amyloid has been shown to be toxic to mature neurons, both in culture and in vivo. Yankner, B. A. et al. (1989) Science 245:417-420; Yankner, B. A. el al. (1990) Proc. Natl. Acad. Sci. USA 87:9020-9023; Roher, A. E. et al. (1991) Biochem. Biophys. Res. Commun. 174:572-579; Kowall, N. W. et al. (1991) Proc. Natl. Acad. Sci. USA 88:7247-7251. Furthermore, patients with hereditary cerebral hemorrhage with amyloidosis-Dutch-type (HCHWA-D), which is characterized by diffuse .beta.-amyloid deposits within the cerebral cortex and cerebrovasculature, have been shown to have a point mutation that leads to an amino acid substitution within .beta.-AP. Levy, E. et al. (1990) Science 248:1124-1126. This observation demonstrates that a specific alteration of the .beta.-AP sequence can cause .beta.-amyloid to be deposited.
Natural .beta.-AP is derived by proteolysis from a much larger protein called the amyloid precursor protein (APP). Kang, J. et al. (1987) Nature 325:733; Goldgaber, D. et al. (1987) Science 235:877; Robakis, N. K. et al. (1987) Proc. Natl. Acad. Sci. USA 84:4190; Tanzi, R. E. et al. (1987) Science 235:880. The APP gene maps to chromosome 21, thereby providing an explanation for the .beta.-amyloid deposition seen at an early age in individuals with Down's syndrome, which is caused by trisomy of chromosome 21. Mann, D. M. et al. (1989) Neuropathol. Appl. Neurobiol. 15:317; Rumble, B. et al. (1989) N. Eng. J. Med. 320:1446. APP contains a single membrane spanning domain, with a long amino terminal region (about two-thirds of the protein) extending into the extracellular environment and a shorter carboxy-terminal region projecting into the cytoplasm. Differential splicing of the APP messenger RNA leads to at least five forms of APP, composed of either 563 amino acids (APP-563), 695 amino acids (APP-695), 714 amino acids (APP-714), 751 amino acids (APP-751) or 770 amino acids (APP-770).
Within APP, naturally-occurring .beta. amyloid peptide begins at an aspartic acid residue at amino acid position 672 of APP-770. Naturally-occurring .beta.-AP derived from proteolysis of APP is 39 to 43 amino acid residues in length, depending on the carboxy-terminal end point, which exhibits heterogeneity. The predominant circulating form of .beta.-AP in the blood and cerebrospinal fluid of both AD patients and normal adults is .beta.1-40 ("short .beta."). Seubert, P. et al. (1992) Nature 359:325; Shoji, M. et al. (1992) Science 258:126. However, .beta.1-42 and .beta.1-43 ("long .beta.") also are forms in .beta.-amyloid plaques. Masters, C. et al. (1985) Proc. Natl. Acad. Sci. USA 82:4245; Miller, D. et al. (1993) Arch. Biochem. Biophys. 301:41; Mori, H. et al. (1992) J. Biol. Chem. 267:17082. Although the precise molecular mechanism leading to .beta.-APP aggregation and deposition is unknown, the process has been likened to that of nucleation-dependent polymerizations, such as protein crystallization, microtubule formation and actin polymerization. See e.g., Jarrett, J. T. and Lansbury, P. T. (1993) Cell 73:1055-1058. In such processes, polymerization of monomer components does not occur until nucleus formation. Thus, these processes are characterized by a lag time before aggregation occurs, followed by rapid polymerization after nucleation. Nucleation can be accelerated by the addition of a "seed" or preformed nucleus, which results in rapid polymerization. The long .beta. forms of .beta.-AP have been shown to act as seeds, thereby accelerating polymerization of both long and short .beta.-AP forms. Jarrett, J. T. et al. (1993) Biochemistry 32:4693.
In one study, in which amino acid substitutions were made in .beta.-AP, two mutant .beta. peptides were reported to interfere with polymerization of non-mutated .beta.-AP when the mutant and non-mutant forms of peptide were mixed. Hilbich, C. et al. (1992) J. Mol. Biol. 228:460-473. Equimolar amounts of the mutant and non-mutant (i.e., natural) .beta. amyloid peptides were used to see this effect and the mutant peptides were reported to be unsuitable for use in vivo. Hilbich, C. et al. (1992), supra.